- Email: [email protected]
Magnetoelectrical stimulation of motor cortex in children with motor disturbances
Electroencephalography and clinical Neurophysiology, 85 (1992) 86-94 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0924-980X/92/$05.00
86
ELMOCO 90666
M a g n e t o e l e c t r i c a l s t i m u l a t i o n o f m o t o r cortex in c h i l d r e n with m o t o r d i s t u r b a n c e s * K. Miiller
a,
V. H6mberg b, A. Aulich c and H.G. Lenard
a
a Department of Pediatrics, b Neurological Therapy Center and c Department of Neurology, Heinrich-Heine-University, Diisseldorf (F.R.G.) (Accepted for publication: 5 November 1991)
Summary Transcranial magnetoelectrical stimulation (TMS) is now widely used as a diagnostic tool in adults. In this study we report our experiences with this technique in children with central motor disturbances. We used a Cadwell MES10 magnetoelectrical stimulator with a maximal magnetic field of 2 tesla. The stimulation procedure followed a standardized protocol, with the patients being as relaxed as possible in order to avoid contamination of parameters with different preinnervational levels. Stimulation data were compared to a data base obtained in 58 normal children. The first group of patients consisted of 20 children aged from 7 months to 16 years with hemiparesis of different etiologies. Neuroimaging data were correlated with the results of magnetoelectrical stimulation. In 13 patients a pathological pattern of TMS could be detected, and in 7 of these a corresponding lesion of the cortico-spinal tract was found in CT or MRI scans. In 7 children TMS was normal, in spite of a clear-cut lesion of the cortico-spinal tract in CT or MRI scans in 4 of them. The second group of patients consisted of 16 children with extrapyramidal disease, mostly of hereditary origin, such as DOPA-responsive dystonia or benign hereditary chorea. TMS showed a normal response pattern in this group. We discuss problems and possible pitfalls in TMS in childhood in evaluating the diagnostic value of TMS. At the moment the diagnostic usefulness of TMS in children with motor disturbances appears limited and calls for careful interpretation.
Key words" Transcranial cortical stimulation; Cortico-spinal tract; Hemiparesis; Upper motor neuron syndrome; Extrapyramidal disease
Transcranial magnetoelectrical stimulation (TMS) of motor cortex and peripheral nerve roots (Barker et al. 1985) is now widely used in adults to describe pathology of motor efferents in a variety of motor disorders (e.g., Ingram and Swash 1987; Hess et al. 1987; Claus et al. 1988; Ingram et al. 1988). Compared to transcranial electrical stimulation TMS has the advantage of being almost painless and hence can also be used as a non-invasive tool in children. Animal experimental evidence suggests that TMS can be used safely (e.g., Agnew and McCreery 1987; Eyre et al. 1990). Normative data regarding the maturation of the fastest cortico-spinal efferents have already been collected (Koh and Eyre 1988; Miiller et al. 1991). In children with motor disturbances it may be difficult clinically to determine which part of the motor system gives rise to
Correspondence to: Dr. K. Miiller, Department of Pediatrics, University of Diisseldorf, Moorenstrasse 5, D-4000 Diisseldorf 1 (F.R.G.). * This work was supported by a grant from the Ministry of Science and Technology of the Federal Republic of Germany (0706560).
the observed signs and symptoms, because brain lesions are often global, e.g., as in peri- or postnatal hypoxia. In such cases TMS may provide additional information about the relative contribution of the cortico-spinal tract in comparison to "extrapyramidal" pathways in the generation of motor disorders. In patients with unilateral infarctions both diminution of amplitude and prolongation of latency of responses contralateral to the affected hemisphere have been demonstrated with electrical stimulation or TMS (Berardelli et al. 1987; Macdonnel et al. 1989; H6mberg et al. 1991). In contrast, in children and adults with degenerative diseases of the basal ganglia, responses after electrical or magnetoelectrical stimulation of motor cortex are normal (Dick et al. 1984; Thompson et al. 1986; Miiller et al. 1989a,b; H/Smberg and Lange 1990). The goal of the present study was to examine the relationship between TMS and the underlying morphological substrate determined by neuroimaging in children presenting with either a unilateral upper motor neuron syndrome or extrapyramidal signs and symptoms.
TMS OF MOTOR CORTEX IN CHILDREN
87
Methods Neuroimaging The neuroimaging data (MRI or CT scans) regarding the cortico-spinal tract were assessed by a neuroradiologist (A.A.) according to the following criteria: focal lesions of the cortico-spinal system were divided into cortical defects in motor or premotor areas and subcortical lesions (detectable defects in the corona radiata, internal capsule or brain-stem).
Patients The group with unilateral upper motor neuron syndrome (UMN group) consisted of 20 children aged between 7 months and 16 years (13 males, 7 females). They presented with a clearly defined hemiparesis, variable in clinical severity and caused by different etiologies. The second group (EPMS group) consisted of 16 patients aged from 4.5 to 16 years (12 males, 4 females) with hereditary or acquired forms of basal ganglia disorder, such as benign hereditary chorea (N = 4), Gilles de la Tourette's syndrome (N = 4), DOPA-responsive dystonia (N--3) or dystonia or chorea of unknown origin (N =4). One patient had had a hemidystonia since birth and showed a circumscribed lesion in the basal ganglia and in the cortico-spinal tract in the CT scan.
Data recording and analysis A Cadwell MES10 magnetoelectrical stimulator was used with a maximum magnetic field of 2 tesla pulsed for 100/zsec with a bipolar pulse shape of the induced electrical field. A standardized examination procedure was carried out, with the children lying supine and being as relaxed as possible. It is known from adults that even slight differences in preinnervation state can considerably affect response latency (e.g., Day et al. 1986; Hess et al. 1986). This is also true in children. An example is illustrated in Fig. 1. This shows responses at
2o.0 ~ I~
right t h e n a r
~ , I
I 200 pV I
lOreS
Fig. 1. TMS displays 4 different response latencies at the right thenar due to different preinnervational levels in this 3-year-old girl with cerebral palsy, who was unable to remain relaxed during stimulation.
the right thenar in a mentally retarded 3-year-old child with cerebral palsy, who presented with a marked "spasticity" of all 4 extremities and athetotic movements in the upper extremities. In this child, who was neither able to stay quiet nor to follow instructions, different levels of preinnervation gave rise to a change in latency at the right thenar by up to 8 msec. Nevertheless, even the longest latency remained in the normal age range. This demonstrates the difficulty in interpreting latencies unless the examination is restricted to a relaxed state. In our experience latency variability even in normal children is much more marked than in adults. EMG surface electrodes were attached bilaterally over the bellies of m. abductor poltieis brevis at the thenar eminence and, for the lower extremities, at the m. abductor hallucis. The stimulator coil (diameter 9.5 cm) was positioned over the vertex region. Stimulation intensity was gradually increasecl.in steps of 10%, starting at an intensity of 40% of maximum stimulator output, in order to determine stimulation thresholds. Definite measurements of latencies and amplitudes of EMG responses were always obtained at the full 2.0 tesla intensity. The full details of the procedure have been described in a preceding study (Miiller et al. 1991). To determine central conduction times (CCTs), cervical and lumbar root stimulations were performed in addition to cortical stimulation by positioning the magnetic coil in the midline over the lower cervical and lower lumbar vertebrae, respect~ely. The latency difference between cortical and root stimulation was used to estimate CCTs. EMG responses were obtained using a computerized 4-channel EMG recording system with a band pass flat between 10 Hz and 2 kHz. All responses were stored on magnetic tape for off-line analysis. The patients' data were compared to a normative data base, obtained in 58 children between the ages of 1 year and 13 years, using the same stimulation protocol (Miiller et al. 1991). This study had demonstrated that, in relaxed muscles, responses in the upper extremity (thenar) were reproducibly obtainable after the age of 15 months, whereas responses in the lower extremity could not be reliably obtained before the fifth year of life. Pathological features of the obtained EMG responses after cortical stimulation were defined as follows. (1) Absolute CCT measures exceeding 2.5 S.D.s of age-related norms. (2) Right.left CCT differences of more than 1.5 msec for the upper extremity and 2.1 msec for the lower extremity (i.e., 2.5 S.D,s of right-left side differences in the normal control group) when absolute CCTs on the affected side were still in the normal range.
88
(3) Complete absence of an evoked response after cortical stimulation in the relaxed state on the affected side in children older than 15 months for the upper, and older than 6 years for the lower extremity. These boundaries were defined by the fact that unequivocal responses can be obtained in normal children beyond these age limits (MiiUer et al. 1991). (4) The interpretation of absolute amplitude levels
K. MIJLLER ET AL.
or amplitude side differences in children with motor disturbances was difficult, because in normals also absolute amplitude values show a high variability (see Miiller et al. 1991). This is also the case in adults (e.g., Rothwell et al. 1987; HSmberg et al. 1991). Therefore any conclusions based solely on absolute amplitude differences were avoided.
Fig. 2. A: MRI scan of a boy, aged 4 years and 11 months, with a left parietal cystic tumor(for details see Results). B: TMS shows a significant latency difference between the affected and non-affected sides for the upper extremity; in the lower extremity this side difference is only marginal.
89
TMS OF MOTOR CORTEX IN CHILDREN Results
UMN group According to lesion localization by neuroimaging and the outcome of TMS, patient data were subdivided into 4 groups: (I) Clinical presentation corresponded to a focal lesion in the cortico-spinal tract as defined above and TMS showed a pathological pattern. (II) Patients with clinical hemiparesis and pathological TMS but without focal lesions of the cortico-spinal tract on CT or MRI scan. (III) Normal TMS data despite evidence of a focal lesion in the cortico-spinal tract. (IV) Both neuroimaging data and TMS were without evidence of cortico-spinal tract involvement, despite clinical appearance of hemiparesis. The pattern of frequency distribution of TMS and neuroimaging abnormalities is summarized in Table I. UMN I group consisted of 7 children. Fig. 2 shows an example of a 4-year-old child, presenting clinically with only mild paresis of the right arm and leg. The CT scan in the frontal plane shows a cystic tumor of large extent, encompassing premotor and motor cortical areas, as well as the internal capsule and parts of the striatum. TMS reveals a marked latency difference for the upper extremity between the affected right and the unaffected left sides. This is also reflected in prolonged CCTs to the right thenar. Responses of the lower extremity show only a marginal side difference, not considered to be significant. Group UMN II comprised 6 children. Fig. 3 shows CT scan and magnetic stimulation data of a 5.5-year-old male, suffering from a left hemiparesis following global
TABLE I Results of TMS in 20 hemiparetic patients with or without focal lesions of the pyramidal tract (PT) detectable by neuroimaging. Focal lesions of the PT in CT/MRI scan
Abnormal TMS NormalTMS Total UMN I UMN III 11 N= 7 N= 4
No focal lesions of the PT UMN II in CT/MRI scan N= 6
UMN IV N=3
hypoxia after operation for pulmonary stenosis. This was initially reflected in bilateral cerebral white matter hypodensities in a CT scan carried out immediately after the operation. The first TMS study was performed 2 years and 8 months after the operation when, except for a small right frontal and occipital deficit, the CT scan was normal. Magnetic stimulation revealed a latency prolongation to both thenar muscles and a side difference with a longer latency_ to the clinically more affected left thenar compared to the right. At this age, there were no recordable responses from either lower extremity. A second recording performed 14 months later revealed a clear response in the right abductor hallucis with a prolonged central conduction time, but a missing response on the affected left side. This indicates that at a point in time when focal abnormalities were no longer detectable, cortical stimulation still revealed that (due to the global nature of the lesion) cortico-spinal efferents were affected also on the side appearing normal clinically. • Group U M N III represents patients with clinical hemiparesis and a concomitant focal lesion in neu-
B
I 2oo pY thenar
I
34.4
9
I 20o pY
I so pV
,., ~ ^
abd.hall. 5opY
Fig. 2 (continued).
90
roimaging, in whom responses to cortical stimulation were normal (N = 4). Fig. 4 shows the example of a boy aged 15 years and 3 months at the time of the TMS study. The CT reveals a large left fronto-precentral lesion, certainly affecting motor and premotor cortical areas. TMS data, however, appear normal. In another case (Fig. 5) TMS appeared normal without significant side differences. Responses of the affected right extremities appeared shorter in latency and larger in amplitude than those of the non-affected left side. This was caused by overlying tonic spastic muscular activity on the affected side. This illustrates another possible pitfall in the interpretation of TMS responses in cases where relaxation is insufficient. Group UMN IV was characterized by a clinical hemiparetic syndrome with neither neuroimaging nor TMS revealing focal abnormalities of cortico-spinal efferents (N = 3). Fig. 6 illustrates data of a 16-year-01d female with a mild right-sided hemiparesis, due to a
K. M U L L E R E T AL.
connatal left-sided parieto-occipital porencephalic cyst on CT scan. Magnetic stimulation data show normal symmetrical latencies. In summary, 13 out of 20 patients with clinical UMN syndrome showed a pathological pattern of responses after magnetoelectrical motor cortex stimulation. In 7 of these children a corresponding lesion of the motor cortex or its descending fibers was detected on MRI or CT scans. In the other 6 patients, despite parallel clinical and electrophysiological evidence of a corticospinal lesion, neuroimaging did not reveal focal abnormalities but showed generalized damage such as cortical atrophy or hydrocephalus in some cases. This indicates that in these cases cortical stimulation was more sensitive in demonstrating a !cortico-spinai lesion than was neuroimaging. In all of these 13 patients the patterns of abnormalities of responses after Cortical stimulation were similar, with a predominant latency prolongation on the af-
J~Jl 23.6
.5,6 yeor. .
26.0
i,oo.v
'
11
.
.
.
.
Ih OAldF
II
~d.hll
--
t
loM8
. ~, 21.0 , # j ~ , i t
6 8 years
I",~
t~nar
,/i,:
|0
Ill'
46,0 -
.
~_
---,
,~ t...;:'.-
Ills
[
=.
SO
pV
aIXI.I'MII 1.
!
50 IJ¥
10 IB
B Fig. 3. A: the C T scan of a boy, 6 years and 8 m o n t h s old, with left hemiparesis shows a right frontal and occipital cortical defect. B: stimulation at the age of 5.5 years reveals prolonged latencies at both left and right thenar muscles with a side difference on the hemiparetic left side. There are no responses in the lower extremities, which can nevertheless be normal at this age. At the age of 6 years and 8 m o n t h s T M S still shows the side difference for the upper extremity as well as a prolonged response on the clinically non-affected side. For the lower extremity the response is missing on the clinically affected left side.
TMS OF MOTOR CORTEX IN CHILDREN
_ 20.7
------'--
j
91
_ ~ : ~
F
r
thenar
I
I I
~ ,S., J
U
v
,ba.hal~. I
~ . - - ~
1500/Jv lOms
fected side. This was usually accompanied by a marked amplitude diminution of responses on the affected side up to completely missing responses during relaxation. In 3 patients with clinically apparent hemiparetic syndromes, neuroimaging and TMS were concordant in not revealing abnormalities of cortico-spinal descending fibers, which may indicate that primarily extrapyramidal pathology, or lesions not detectable by these examination techniques, was responsible for the clinical picture. Finally, in 4 hemiparetic patients, despite neuroradiological evidence of lesions affecting the cortico-spinal tract, cortical stimulation failed to reflect these lesions.
Fig. 4. A: CT scan of a male, 15 years and 3 months old, showing a left froXitalprecentral lesion, affecting areas 4, 6 and SMA extending deep inio the white matter, the caudate and the thalamus, as well as cortical atrophy with enlargement of the left lateral ventricle. B: the corresponding response pattern to TMS is normal. E P M S group In the EPMS group CT or MRI scans of 15 patients were normal. In all of these children magnetic stimulation revealed normal results. This was also the case in a patient with connatal hemJdystonia, who showed an enlarged left lateral ventricle, a large hypodensity in the left hypothalamic and thalamic region and internal capsule but had normal central conduction times.
Discussion In the majority of children (13 of 20) with UMN syndromes abnormalities of responses after TMS were
K. MULLER ET AL.
92
16.0,~
~
,
v
then!r
I loopV abd.hall.
1o0pV
33.2m ~---,-~- -
-~ -
-,
~
~.
.
r~- . -~- ~-= - --~
....
~.... .
.
.
- -
-~ - ~ - ~ -~
~
\
lorns Fig. 5. Despite fight hemiparesis, TMS of a boy, 7 years and 5 months old, results in normal symmetrical latencics. There is, however, overlying involuntary "spastic" muscle activity on the affected side.
r
11000 pV
thenar
1 I,o00,V 10
f
I.
":~
.....
ms
_
z..~
500
abd. ha I I.
3.,
A "
,
t~J
Is00
,o.s
Fig. 6. A: the CT scan of a 16-year-old female with a mild right hemiparesis shows a connatal left parieto occipital porencephalic cyst and hemiatrophy of the left hemisphere, sparing the motor cortex. B: the corresponding TMS produces normal responses.
°v
TMS OF MOTOR CORTEX IN CHILDREN
obtained. These appeared primarily in the form of prolonged latencies between motor cortex and the affected muscle, or prolonged CCTs between motor cortex and spinal motor neuron pools, compared to the non-affected side. Absolute prolongation of CCTs or latencies were much less frequent, due to the wide range of normal data in each age group. This resembles the pattern found also in adults using electrical (Berardelli et al. 1987; Thompson et al. 1987) or magnetoelectrical stimulation (Macdonnel et al. 1989; H6mberg et al. 1991) after ischemic hemispheric lesions. In adults with large cortical lesions a complete loss of responses has also been observed. Due to higher stimulation thresholds in normal children compared to adults, interpretation of missing responses is difficult. In contrast to the situation in adults, a missing response in children may just reflect threshold elevation rather than a structural lesion. On the other hand we also found decreased thresholds on the affected side in "spastic" patients, in whom responses could be evoked in the affected leg. Hence both threshold elevation and decrease, compared to the non-affected side, may be the consequence of a central lesion. The sensitivity of TMS in picking up pathology not present on neuroimaging is demonstrated in the examples of group UMN II: in several children the electrophysiological approach seemed to be more sensitive than imaging in detecting cortico-spinal tract abnormalities, corresponding to the clinical picture. This shows that TMS is at least useful as an adjunct to neuroimaging in the assessment of children with unilateral motor disturbances. Another important finding from this study is that children with extrapyramidal pathology, such as DOPA-responsive dystonia, benign hereditary chorea, writers' cramp or Gilles de la Tourette syndrome, show invariably normal TMS responses despite clear motor "plus" signs (hyperkinesia, tremor) and motor deficits in the form of bradykinesia. This parallels findings obtained in adult patients with degenerative basal ganglia disorders, such as Parkinson's disease (Dick et al. 1984; Thompson et al. 1986) or Huntington's disease (H6mberg and Lange 1990). The potential of the method to differentiate between "pyramidal" and "extrapyramidal" lesions might be helpful in delineating the affected structures more precisely in children with "cerebral palsy." Up to now, TMS of motor cortex in children has provided insight into the maturation of cortico-spinal descending fibers in normals (Koh and Eyre 1988; Miiller et al. 1991). The present study shows that TMS may also provide useful information for the differential diagnosis of motor disturbances in children. TMS results, however, have to be treated and interpreted more carefully in children than in adults, considering the wide variability of normal data dependent on dif-
93
ferent preinnervation levels and the additional factor of age-dependent changes in stimulation thresholds. This could limit the application of TMS for exact quantitative analysis of the cortico-spinal system in children with neurological problems. Especially in younger children, i.e., the age of 5-6 years, where thresholds play an important role, the method appears to be less reliable than in adults.
References Agnew, W.F. and McCreery, D.B. Considerations for safety in the use of extracranial stimulation for motor evoked potentials. Neurosurgery, 1987, 20: 143-147. Barker, A.T., Freeston, I.L., Jalinous, R., Merton, P.A. and Morton, H.B. Magnetic stimulation of the human brain. J. Physiol. (Lond.), 1985, 369: 3P. Berardelli, A., Inghilleri, M., Manfredi, M., Zamponi, A., Cecconi, V, and Dolce, G. Cortical and cervical stimulation after hemispheric infarction. J. Neurol. Neurosurg. Psychiat., 1987, 50: 861-865. Claus, D., Harding, A.E., Hess, C.W., Mills, K.R. and Murray, N.M.F. Central motor conduction in degenerative ataxic disorders: a magnetic stimulation study. J. Neurol. Neurosurg. Psychiat., 1988, 51: 790-795. Day, B.L., Dick, J.P.R., Marsden, C.D. and Thompson, P.D. Differences between electrical and magnetic stimulation of the human brain. J. Physiol. (Lond.), 1986, 378: 36P. Dick, J.P.R., Cowan, M.A., Day, B.L., Berardelli, A., Kachi, T., Rothwell, J.C. and Marsden, C.D. The corticomotoneuron connection is normal in Parkinson's disease. Nature, 1984, 310: 407-409. Eyre, J.A., Flecknell, P.A., Kenyon, B.R., Koh, T.H.H.G. and Miller, S. Acute effects of electromagnetic stimulation of the brain on cortical activity, cortical blood flow, blood pressure and heart rate in the cat: an evaluation of safety. J. Neurol. Neurosurg. Psychiat., 1990, 53: 507-513. Hess, C.W., Mills, K,R. and Murray, N.M.F. Magnetic stimulation of the human brain: the effects of voluntary muscle activity. J. Physiol. (Lond.), 1986, 378: 37P. Hess, C.W., Mills, K.R., Murray, N.M.F. and Schriefer, T.N. Magnetic brain stimulation: central motor conduction studies in multiple sclerosis. Ann. Neurol., 1987, 22: 744-752. H6mberg, V. and Lange, H. Central motor conduction to hand and leg muscles in Huntington's disease. Movem. Dis., 1990, 5: 214218. H6mberg, V., Stephan, K.M. and Netz, J. Transcranial stimulation of motor cortex in upper motor neurone syndrome: its relation to the motor deficit. Electroenceph. clin. Neurophysiol., 1991, 81: 377-388. Ingram, D.A. and Swash, M. Central motor conduction is abnormal in motor neuron disease. J. Neurol. Neurosurg. Psychiat., 1987, 50: 159-166. Ingram, D.A., Thompson, A.J. and Swash, M. Central motor conduction in multiple sclerosis: evaluation of abnormalities revealed by transcutaneous magnetic stimulation of the brain. J. Neurol. Neurosurg. Psychiat., 1988, 51: 487-494. Koh, T.H.H.G. and Eyre, J.A. Maturation of cortico-spinal tracts assessed by electromagnetic stimulation of the motor cortex. Arch. Dis. Child., 1988, 63: 1347-1352. Macdonnel, R.A.L., Donnan, A. and Bladin, P.F. A comparison of somatosensory evoked and motor evoked potentials in stroke. Ann. Neurol., 1989, 25: 68-73.
94 Miiller, K., H6mberg, V. and Lenard,/-I.G. Motor control in childhood onset DOPA-responsive dystonia (Segawa syndrome). Neuropaediatrics, 1989a, 20: 185-191. Miiller, IC, H6mberg, V. and Lenard, H.G. Benigne heredit~ire Chorea: klinisch neurophysiologische Untersuchungen. in: H.M. Weinmann (Ed.), Aktuelle Neurop~idiatrie 88. Springer, Berlin, 1989b: 188-192. Miiller, K., H6mberg, V. and Lenard, H.G. Magnetoelectrical stimulation of motor cortex and nerve roots in children. Maturation of cortico-motoneural projections. Electroenceph. clin. Neurophysio|., 1991, 81: 63-70. Rothwell, J.C., Thompson, P.D., Day, B.L., Dick, J.P.R., Kachi, T., Cowan, J.M.A. and Marsden, C.D. Motor cortex stimulation in
K. Mf0LLER ET AL. intact man. I. General characteristics of EMG responses in different muscles. Brain, 1987, 110: 1173-1190. Thompson, P.D., Dick, J.P.R., Day, B.L., Rothwell, J.C., Berardelli, A., Kachi, T. and Marsden, C.D. Electrophysiology of the cortico-motoneurone pathways in patients with movement disorders. Movem. Dis., 1986, 2: 113-117. Thompson, P.D., Day, B.L., Rothwell, J.C., Dick, J.P.R., Cowan, J.M.A., Asselman, P., Griffin, G.B., Sheely, M.P. and Marsden, C.D. The interpretation of electromyographic responses to electrical stimulation of the motor cortex in disease of upper motor neurone. J. Neurol. Sci., 1987, 80: 91-110.
86
ELMOCO 90666
M a g n e t o e l e c t r i c a l s t i m u l a t i o n o f m o t o r cortex in c h i l d r e n with m o t o r d i s t u r b a n c e s * K. Miiller
a,
V. H6mberg b, A. Aulich c and H.G. Lenard
a
a Department of Pediatrics, b Neurological Therapy Center and c Department of Neurology, Heinrich-Heine-University, Diisseldorf (F.R.G.) (Accepted for publication: 5 November 1991)
Summary Transcranial magnetoelectrical stimulation (TMS) is now widely used as a diagnostic tool in adults. In this study we report our experiences with this technique in children with central motor disturbances. We used a Cadwell MES10 magnetoelectrical stimulator with a maximal magnetic field of 2 tesla. The stimulation procedure followed a standardized protocol, with the patients being as relaxed as possible in order to avoid contamination of parameters with different preinnervational levels. Stimulation data were compared to a data base obtained in 58 normal children. The first group of patients consisted of 20 children aged from 7 months to 16 years with hemiparesis of different etiologies. Neuroimaging data were correlated with the results of magnetoelectrical stimulation. In 13 patients a pathological pattern of TMS could be detected, and in 7 of these a corresponding lesion of the cortico-spinal tract was found in CT or MRI scans. In 7 children TMS was normal, in spite of a clear-cut lesion of the cortico-spinal tract in CT or MRI scans in 4 of them. The second group of patients consisted of 16 children with extrapyramidal disease, mostly of hereditary origin, such as DOPA-responsive dystonia or benign hereditary chorea. TMS showed a normal response pattern in this group. We discuss problems and possible pitfalls in TMS in childhood in evaluating the diagnostic value of TMS. At the moment the diagnostic usefulness of TMS in children with motor disturbances appears limited and calls for careful interpretation.
Key words" Transcranial cortical stimulation; Cortico-spinal tract; Hemiparesis; Upper motor neuron syndrome; Extrapyramidal disease
Transcranial magnetoelectrical stimulation (TMS) of motor cortex and peripheral nerve roots (Barker et al. 1985) is now widely used in adults to describe pathology of motor efferents in a variety of motor disorders (e.g., Ingram and Swash 1987; Hess et al. 1987; Claus et al. 1988; Ingram et al. 1988). Compared to transcranial electrical stimulation TMS has the advantage of being almost painless and hence can also be used as a non-invasive tool in children. Animal experimental evidence suggests that TMS can be used safely (e.g., Agnew and McCreery 1987; Eyre et al. 1990). Normative data regarding the maturation of the fastest cortico-spinal efferents have already been collected (Koh and Eyre 1988; Miiller et al. 1991). In children with motor disturbances it may be difficult clinically to determine which part of the motor system gives rise to
Correspondence to: Dr. K. Miiller, Department of Pediatrics, University of Diisseldorf, Moorenstrasse 5, D-4000 Diisseldorf 1 (F.R.G.). * This work was supported by a grant from the Ministry of Science and Technology of the Federal Republic of Germany (0706560).
the observed signs and symptoms, because brain lesions are often global, e.g., as in peri- or postnatal hypoxia. In such cases TMS may provide additional information about the relative contribution of the cortico-spinal tract in comparison to "extrapyramidal" pathways in the generation of motor disorders. In patients with unilateral infarctions both diminution of amplitude and prolongation of latency of responses contralateral to the affected hemisphere have been demonstrated with electrical stimulation or TMS (Berardelli et al. 1987; Macdonnel et al. 1989; H6mberg et al. 1991). In contrast, in children and adults with degenerative diseases of the basal ganglia, responses after electrical or magnetoelectrical stimulation of motor cortex are normal (Dick et al. 1984; Thompson et al. 1986; Miiller et al. 1989a,b; H/Smberg and Lange 1990). The goal of the present study was to examine the relationship between TMS and the underlying morphological substrate determined by neuroimaging in children presenting with either a unilateral upper motor neuron syndrome or extrapyramidal signs and symptoms.
TMS OF MOTOR CORTEX IN CHILDREN
87
Methods Neuroimaging The neuroimaging data (MRI or CT scans) regarding the cortico-spinal tract were assessed by a neuroradiologist (A.A.) according to the following criteria: focal lesions of the cortico-spinal system were divided into cortical defects in motor or premotor areas and subcortical lesions (detectable defects in the corona radiata, internal capsule or brain-stem).
Patients The group with unilateral upper motor neuron syndrome (UMN group) consisted of 20 children aged between 7 months and 16 years (13 males, 7 females). They presented with a clearly defined hemiparesis, variable in clinical severity and caused by different etiologies. The second group (EPMS group) consisted of 16 patients aged from 4.5 to 16 years (12 males, 4 females) with hereditary or acquired forms of basal ganglia disorder, such as benign hereditary chorea (N = 4), Gilles de la Tourette's syndrome (N = 4), DOPA-responsive dystonia (N--3) or dystonia or chorea of unknown origin (N =4). One patient had had a hemidystonia since birth and showed a circumscribed lesion in the basal ganglia and in the cortico-spinal tract in the CT scan.
Data recording and analysis A Cadwell MES10 magnetoelectrical stimulator was used with a maximum magnetic field of 2 tesla pulsed for 100/zsec with a bipolar pulse shape of the induced electrical field. A standardized examination procedure was carried out, with the children lying supine and being as relaxed as possible. It is known from adults that even slight differences in preinnervation state can considerably affect response latency (e.g., Day et al. 1986; Hess et al. 1986). This is also true in children. An example is illustrated in Fig. 1. This shows responses at
2o.0 ~ I~
right t h e n a r
~ , I
I 200 pV I
lOreS
Fig. 1. TMS displays 4 different response latencies at the right thenar due to different preinnervational levels in this 3-year-old girl with cerebral palsy, who was unable to remain relaxed during stimulation.
the right thenar in a mentally retarded 3-year-old child with cerebral palsy, who presented with a marked "spasticity" of all 4 extremities and athetotic movements in the upper extremities. In this child, who was neither able to stay quiet nor to follow instructions, different levels of preinnervation gave rise to a change in latency at the right thenar by up to 8 msec. Nevertheless, even the longest latency remained in the normal age range. This demonstrates the difficulty in interpreting latencies unless the examination is restricted to a relaxed state. In our experience latency variability even in normal children is much more marked than in adults. EMG surface electrodes were attached bilaterally over the bellies of m. abductor poltieis brevis at the thenar eminence and, for the lower extremities, at the m. abductor hallucis. The stimulator coil (diameter 9.5 cm) was positioned over the vertex region. Stimulation intensity was gradually increasecl.in steps of 10%, starting at an intensity of 40% of maximum stimulator output, in order to determine stimulation thresholds. Definite measurements of latencies and amplitudes of EMG responses were always obtained at the full 2.0 tesla intensity. The full details of the procedure have been described in a preceding study (Miiller et al. 1991). To determine central conduction times (CCTs), cervical and lumbar root stimulations were performed in addition to cortical stimulation by positioning the magnetic coil in the midline over the lower cervical and lower lumbar vertebrae, respect~ely. The latency difference between cortical and root stimulation was used to estimate CCTs. EMG responses were obtained using a computerized 4-channel EMG recording system with a band pass flat between 10 Hz and 2 kHz. All responses were stored on magnetic tape for off-line analysis. The patients' data were compared to a normative data base, obtained in 58 children between the ages of 1 year and 13 years, using the same stimulation protocol (Miiller et al. 1991). This study had demonstrated that, in relaxed muscles, responses in the upper extremity (thenar) were reproducibly obtainable after the age of 15 months, whereas responses in the lower extremity could not be reliably obtained before the fifth year of life. Pathological features of the obtained EMG responses after cortical stimulation were defined as follows. (1) Absolute CCT measures exceeding 2.5 S.D.s of age-related norms. (2) Right.left CCT differences of more than 1.5 msec for the upper extremity and 2.1 msec for the lower extremity (i.e., 2.5 S.D,s of right-left side differences in the normal control group) when absolute CCTs on the affected side were still in the normal range.
88
(3) Complete absence of an evoked response after cortical stimulation in the relaxed state on the affected side in children older than 15 months for the upper, and older than 6 years for the lower extremity. These boundaries were defined by the fact that unequivocal responses can be obtained in normal children beyond these age limits (MiiUer et al. 1991). (4) The interpretation of absolute amplitude levels
K. MIJLLER ET AL.
or amplitude side differences in children with motor disturbances was difficult, because in normals also absolute amplitude values show a high variability (see Miiller et al. 1991). This is also the case in adults (e.g., Rothwell et al. 1987; HSmberg et al. 1991). Therefore any conclusions based solely on absolute amplitude differences were avoided.
Fig. 2. A: MRI scan of a boy, aged 4 years and 11 months, with a left parietal cystic tumor(for details see Results). B: TMS shows a significant latency difference between the affected and non-affected sides for the upper extremity; in the lower extremity this side difference is only marginal.
89
TMS OF MOTOR CORTEX IN CHILDREN Results
UMN group According to lesion localization by neuroimaging and the outcome of TMS, patient data were subdivided into 4 groups: (I) Clinical presentation corresponded to a focal lesion in the cortico-spinal tract as defined above and TMS showed a pathological pattern. (II) Patients with clinical hemiparesis and pathological TMS but without focal lesions of the cortico-spinal tract on CT or MRI scan. (III) Normal TMS data despite evidence of a focal lesion in the cortico-spinal tract. (IV) Both neuroimaging data and TMS were without evidence of cortico-spinal tract involvement, despite clinical appearance of hemiparesis. The pattern of frequency distribution of TMS and neuroimaging abnormalities is summarized in Table I. UMN I group consisted of 7 children. Fig. 2 shows an example of a 4-year-old child, presenting clinically with only mild paresis of the right arm and leg. The CT scan in the frontal plane shows a cystic tumor of large extent, encompassing premotor and motor cortical areas, as well as the internal capsule and parts of the striatum. TMS reveals a marked latency difference for the upper extremity between the affected right and the unaffected left sides. This is also reflected in prolonged CCTs to the right thenar. Responses of the lower extremity show only a marginal side difference, not considered to be significant. Group UMN II comprised 6 children. Fig. 3 shows CT scan and magnetic stimulation data of a 5.5-year-old male, suffering from a left hemiparesis following global
TABLE I Results of TMS in 20 hemiparetic patients with or without focal lesions of the pyramidal tract (PT) detectable by neuroimaging. Focal lesions of the PT in CT/MRI scan
Abnormal TMS NormalTMS Total UMN I UMN III 11 N= 7 N= 4
No focal lesions of the PT UMN II in CT/MRI scan N= 6
UMN IV N=3
hypoxia after operation for pulmonary stenosis. This was initially reflected in bilateral cerebral white matter hypodensities in a CT scan carried out immediately after the operation. The first TMS study was performed 2 years and 8 months after the operation when, except for a small right frontal and occipital deficit, the CT scan was normal. Magnetic stimulation revealed a latency prolongation to both thenar muscles and a side difference with a longer latency_ to the clinically more affected left thenar compared to the right. At this age, there were no recordable responses from either lower extremity. A second recording performed 14 months later revealed a clear response in the right abductor hallucis with a prolonged central conduction time, but a missing response on the affected left side. This indicates that at a point in time when focal abnormalities were no longer detectable, cortical stimulation still revealed that (due to the global nature of the lesion) cortico-spinal efferents were affected also on the side appearing normal clinically. • Group U M N III represents patients with clinical hemiparesis and a concomitant focal lesion in neu-
B
I 2oo pY thenar
I
34.4
9
I 20o pY
I so pV
,., ~ ^
abd.hall. 5opY
Fig. 2 (continued).
90
roimaging, in whom responses to cortical stimulation were normal (N = 4). Fig. 4 shows the example of a boy aged 15 years and 3 months at the time of the TMS study. The CT reveals a large left fronto-precentral lesion, certainly affecting motor and premotor cortical areas. TMS data, however, appear normal. In another case (Fig. 5) TMS appeared normal without significant side differences. Responses of the affected right extremities appeared shorter in latency and larger in amplitude than those of the non-affected left side. This was caused by overlying tonic spastic muscular activity on the affected side. This illustrates another possible pitfall in the interpretation of TMS responses in cases where relaxation is insufficient. Group UMN IV was characterized by a clinical hemiparetic syndrome with neither neuroimaging nor TMS revealing focal abnormalities of cortico-spinal efferents (N = 3). Fig. 6 illustrates data of a 16-year-01d female with a mild right-sided hemiparesis, due to a
K. M U L L E R E T AL.
connatal left-sided parieto-occipital porencephalic cyst on CT scan. Magnetic stimulation data show normal symmetrical latencies. In summary, 13 out of 20 patients with clinical UMN syndrome showed a pathological pattern of responses after magnetoelectrical motor cortex stimulation. In 7 of these children a corresponding lesion of the motor cortex or its descending fibers was detected on MRI or CT scans. In the other 6 patients, despite parallel clinical and electrophysiological evidence of a corticospinal lesion, neuroimaging did not reveal focal abnormalities but showed generalized damage such as cortical atrophy or hydrocephalus in some cases. This indicates that in these cases cortical stimulation was more sensitive in demonstrating a !cortico-spinai lesion than was neuroimaging. In all of these 13 patients the patterns of abnormalities of responses after Cortical stimulation were similar, with a predominant latency prolongation on the af-
J~Jl 23.6
.5,6 yeor. .
26.0
i,oo.v
'
11
.
.
.
.
Ih OAldF
II
~d.hll
--
t
loM8
. ~, 21.0 , # j ~ , i t
6 8 years
I",~
t~nar
,/i,:
|0
Ill'
46,0 -
.
~_
---,
,~ t...;:'.-
Ills
[
=.
SO
pV
aIXI.I'MII 1.
!
50 IJ¥
10 IB
B Fig. 3. A: the C T scan of a boy, 6 years and 8 m o n t h s old, with left hemiparesis shows a right frontal and occipital cortical defect. B: stimulation at the age of 5.5 years reveals prolonged latencies at both left and right thenar muscles with a side difference on the hemiparetic left side. There are no responses in the lower extremities, which can nevertheless be normal at this age. At the age of 6 years and 8 m o n t h s T M S still shows the side difference for the upper extremity as well as a prolonged response on the clinically non-affected side. For the lower extremity the response is missing on the clinically affected left side.
TMS OF MOTOR CORTEX IN CHILDREN
_ 20.7
------'--
j
91
_ ~ : ~
F
r
thenar
I
I I
~ ,S., J
U
v
,ba.hal~. I
~ . - - ~
1500/Jv lOms
fected side. This was usually accompanied by a marked amplitude diminution of responses on the affected side up to completely missing responses during relaxation. In 3 patients with clinically apparent hemiparetic syndromes, neuroimaging and TMS were concordant in not revealing abnormalities of cortico-spinal descending fibers, which may indicate that primarily extrapyramidal pathology, or lesions not detectable by these examination techniques, was responsible for the clinical picture. Finally, in 4 hemiparetic patients, despite neuroradiological evidence of lesions affecting the cortico-spinal tract, cortical stimulation failed to reflect these lesions.
Fig. 4. A: CT scan of a male, 15 years and 3 months old, showing a left froXitalprecentral lesion, affecting areas 4, 6 and SMA extending deep inio the white matter, the caudate and the thalamus, as well as cortical atrophy with enlargement of the left lateral ventricle. B: the corresponding response pattern to TMS is normal. E P M S group In the EPMS group CT or MRI scans of 15 patients were normal. In all of these children magnetic stimulation revealed normal results. This was also the case in a patient with connatal hemJdystonia, who showed an enlarged left lateral ventricle, a large hypodensity in the left hypothalamic and thalamic region and internal capsule but had normal central conduction times.
Discussion In the majority of children (13 of 20) with UMN syndromes abnormalities of responses after TMS were
K. MULLER ET AL.
92
16.0,~
~
,
v
then!r
I loopV abd.hall.
1o0pV
33.2m ~---,-~- -
-~ -
-,
~
~.
.
r~- . -~- ~-= - --~
....
~.... .
.
.
- -
-~ - ~ - ~ -~
~
\
lorns Fig. 5. Despite fight hemiparesis, TMS of a boy, 7 years and 5 months old, results in normal symmetrical latencics. There is, however, overlying involuntary "spastic" muscle activity on the affected side.
r
11000 pV
thenar
1 I,o00,V 10
f
I.
":~
.....
ms
_
z..~
500
abd. ha I I.
3.,
A "
,
t~J
Is00
,o.s
Fig. 6. A: the CT scan of a 16-year-old female with a mild right hemiparesis shows a connatal left parieto occipital porencephalic cyst and hemiatrophy of the left hemisphere, sparing the motor cortex. B: the corresponding TMS produces normal responses.
°v
TMS OF MOTOR CORTEX IN CHILDREN
obtained. These appeared primarily in the form of prolonged latencies between motor cortex and the affected muscle, or prolonged CCTs between motor cortex and spinal motor neuron pools, compared to the non-affected side. Absolute prolongation of CCTs or latencies were much less frequent, due to the wide range of normal data in each age group. This resembles the pattern found also in adults using electrical (Berardelli et al. 1987; Thompson et al. 1987) or magnetoelectrical stimulation (Macdonnel et al. 1989; H6mberg et al. 1991) after ischemic hemispheric lesions. In adults with large cortical lesions a complete loss of responses has also been observed. Due to higher stimulation thresholds in normal children compared to adults, interpretation of missing responses is difficult. In contrast to the situation in adults, a missing response in children may just reflect threshold elevation rather than a structural lesion. On the other hand we also found decreased thresholds on the affected side in "spastic" patients, in whom responses could be evoked in the affected leg. Hence both threshold elevation and decrease, compared to the non-affected side, may be the consequence of a central lesion. The sensitivity of TMS in picking up pathology not present on neuroimaging is demonstrated in the examples of group UMN II: in several children the electrophysiological approach seemed to be more sensitive than imaging in detecting cortico-spinal tract abnormalities, corresponding to the clinical picture. This shows that TMS is at least useful as an adjunct to neuroimaging in the assessment of children with unilateral motor disturbances. Another important finding from this study is that children with extrapyramidal pathology, such as DOPA-responsive dystonia, benign hereditary chorea, writers' cramp or Gilles de la Tourette syndrome, show invariably normal TMS responses despite clear motor "plus" signs (hyperkinesia, tremor) and motor deficits in the form of bradykinesia. This parallels findings obtained in adult patients with degenerative basal ganglia disorders, such as Parkinson's disease (Dick et al. 1984; Thompson et al. 1986) or Huntington's disease (H6mberg and Lange 1990). The potential of the method to differentiate between "pyramidal" and "extrapyramidal" lesions might be helpful in delineating the affected structures more precisely in children with "cerebral palsy." Up to now, TMS of motor cortex in children has provided insight into the maturation of cortico-spinal descending fibers in normals (Koh and Eyre 1988; Miiller et al. 1991). The present study shows that TMS may also provide useful information for the differential diagnosis of motor disturbances in children. TMS results, however, have to be treated and interpreted more carefully in children than in adults, considering the wide variability of normal data dependent on dif-
93
ferent preinnervation levels and the additional factor of age-dependent changes in stimulation thresholds. This could limit the application of TMS for exact quantitative analysis of the cortico-spinal system in children with neurological problems. Especially in younger children, i.e., the age of 5-6 years, where thresholds play an important role, the method appears to be less reliable than in adults.
References Agnew, W.F. and McCreery, D.B. Considerations for safety in the use of extracranial stimulation for motor evoked potentials. Neurosurgery, 1987, 20: 143-147. Barker, A.T., Freeston, I.L., Jalinous, R., Merton, P.A. and Morton, H.B. Magnetic stimulation of the human brain. J. Physiol. (Lond.), 1985, 369: 3P. Berardelli, A., Inghilleri, M., Manfredi, M., Zamponi, A., Cecconi, V, and Dolce, G. Cortical and cervical stimulation after hemispheric infarction. J. Neurol. Neurosurg. Psychiat., 1987, 50: 861-865. Claus, D., Harding, A.E., Hess, C.W., Mills, K.R. and Murray, N.M.F. Central motor conduction in degenerative ataxic disorders: a magnetic stimulation study. J. Neurol. Neurosurg. Psychiat., 1988, 51: 790-795. Day, B.L., Dick, J.P.R., Marsden, C.D. and Thompson, P.D. Differences between electrical and magnetic stimulation of the human brain. J. Physiol. (Lond.), 1986, 378: 36P. Dick, J.P.R., Cowan, M.A., Day, B.L., Berardelli, A., Kachi, T., Rothwell, J.C. and Marsden, C.D. The corticomotoneuron connection is normal in Parkinson's disease. Nature, 1984, 310: 407-409. Eyre, J.A., Flecknell, P.A., Kenyon, B.R., Koh, T.H.H.G. and Miller, S. Acute effects of electromagnetic stimulation of the brain on cortical activity, cortical blood flow, blood pressure and heart rate in the cat: an evaluation of safety. J. Neurol. Neurosurg. Psychiat., 1990, 53: 507-513. Hess, C.W., Mills, K,R. and Murray, N.M.F. Magnetic stimulation of the human brain: the effects of voluntary muscle activity. J. Physiol. (Lond.), 1986, 378: 37P. Hess, C.W., Mills, K.R., Murray, N.M.F. and Schriefer, T.N. Magnetic brain stimulation: central motor conduction studies in multiple sclerosis. Ann. Neurol., 1987, 22: 744-752. H6mberg, V. and Lange, H. Central motor conduction to hand and leg muscles in Huntington's disease. Movem. Dis., 1990, 5: 214218. H6mberg, V., Stephan, K.M. and Netz, J. Transcranial stimulation of motor cortex in upper motor neurone syndrome: its relation to the motor deficit. Electroenceph. clin. Neurophysiol., 1991, 81: 377-388. Ingram, D.A. and Swash, M. Central motor conduction is abnormal in motor neuron disease. J. Neurol. Neurosurg. Psychiat., 1987, 50: 159-166. Ingram, D.A., Thompson, A.J. and Swash, M. Central motor conduction in multiple sclerosis: evaluation of abnormalities revealed by transcutaneous magnetic stimulation of the brain. J. Neurol. Neurosurg. Psychiat., 1988, 51: 487-494. Koh, T.H.H.G. and Eyre, J.A. Maturation of cortico-spinal tracts assessed by electromagnetic stimulation of the motor cortex. Arch. Dis. Child., 1988, 63: 1347-1352. Macdonnel, R.A.L., Donnan, A. and Bladin, P.F. A comparison of somatosensory evoked and motor evoked potentials in stroke. Ann. Neurol., 1989, 25: 68-73.
94 Miiller, K., H6mberg, V. and Lenard,/-I.G. Motor control in childhood onset DOPA-responsive dystonia (Segawa syndrome). Neuropaediatrics, 1989a, 20: 185-191. Miiller, IC, H6mberg, V. and Lenard, H.G. Benigne heredit~ire Chorea: klinisch neurophysiologische Untersuchungen. in: H.M. Weinmann (Ed.), Aktuelle Neurop~idiatrie 88. Springer, Berlin, 1989b: 188-192. Miiller, K., H6mberg, V. and Lenard, H.G. Magnetoelectrical stimulation of motor cortex and nerve roots in children. Maturation of cortico-motoneural projections. Electroenceph. clin. Neurophysio|., 1991, 81: 63-70. Rothwell, J.C., Thompson, P.D., Day, B.L., Dick, J.P.R., Kachi, T., Cowan, J.M.A. and Marsden, C.D. Motor cortex stimulation in
K. Mf0LLER ET AL. intact man. I. General characteristics of EMG responses in different muscles. Brain, 1987, 110: 1173-1190. Thompson, P.D., Dick, J.P.R., Day, B.L., Rothwell, J.C., Berardelli, A., Kachi, T. and Marsden, C.D. Electrophysiology of the cortico-motoneurone pathways in patients with movement disorders. Movem. Dis., 1986, 2: 113-117. Thompson, P.D., Day, B.L., Rothwell, J.C., Dick, J.P.R., Cowan, J.M.A., Asselman, P., Griffin, G.B., Sheely, M.P. and Marsden, C.D. The interpretation of electromyographic responses to electrical stimulation of the motor cortex in disease of upper motor neurone. J. Neurol. Sci., 1987, 80: 91-110.